Page 1 of 1

Checking multicollinearity&heteroscedasticity for panel data

Posted: Thu May 12, 2011 3:27 am
by christina.sakali
Hi all,

I was wondering how I can check for multicollinearity and heteroscedasticity for Panel Data on EViews. I run a multiple linear regression with FDI as the. dependent variable and 5-6 independent variables on the right hand-side. All variables except a dummy var. are in LOG form and I have 154 observations.

As I had indication for multicollinearity in my data (from the estimation results and the correlation matrix), I decided to divide my variables in two specifications and results now seem to be ok. However I would like to know if there is a more formal way to check for multicollinearity and hereroscedasticity in the case of PANEL data.

I am using EViews 7.

Thank you very much!

Re: Checking multicollinearity&heteroscedasticity for panel

Posted: Fri May 13, 2011 10:35 am
by carmomi
I have the same doubt...